fpls-10-00329 March 14, 2019 Time: 16:28 # 1

ORIGINAL RESEARCHpublished: 18 March 2019

doi: 10.3389/fpls.2019.00329

Edited by:Kazimierz Trebacz,

Maria Curie-Skłodowska University,Poland

Reviewed by:Marina Gavilanes-Ruiz,

National Autonomous Universityof Mexico, Mexico

Jean-luc Cacas,AgroParisTech Institut des Sciences

et Industries du Vivant etde l’Environnement, France

*Correspondence:Magali Deleu

magali.deleu@uliege.be

†Co-last authors

Specialty section:This article was submitted to

Plant Physiology,a section of the journal

Frontiers in Plant Science

Received: 05 December 2018Accepted: 28 February 2019

Published: 18 March 2019

Citation:Lebecque S, Lins L, Dayan FE,

Fauconnier M-L and Deleu M (2019)Interactions Between NaturalHerbicides and Lipid Bilayers

Mimicking the Plant PlasmaMembrane. Front. Plant Sci. 10:329.

doi: 10.3389/fpls.2019.00329

Interactions Between NaturalHerbicides and Lipid BilayersMimicking the Plant PlasmaMembraneSimon Lebecque1,2, Laurence Lins1, Franck E. Dayan3, Marie-Laure Fauconnier4† andMagali Deleu1*†

1 TERRA, Laboratory of Molecular Biophysics at Interfaces, Gembloux Agro-Bio Tech, University of Liège, Gembloux,Belgium, 2 TERRA – AgricultureIsLife, Gembloux Agro-Bio Tech, University of Liège, Gembloux, Belgium, 3 Departmentof Bioagricultural Sciences and Pest Management, Colorado State University, Fort Collins, CO, United States, 4 Generaland Organic Chemistry Laboratory, Gembloux Agro-Bio Tech, University of Liège, Gembloux, Belgium

Natural phytotoxic compounds could become an alternative to traditional herbicides inthe framework of sustainable agriculture. Nonanoic acid, sarmentine and sorgoleoneare such molecules extracted from plants and able to inhibit the growth of variousplant species. However, their mode of action is not fully understood and despite cluesindicating that they could affect the plant plasma membrane, molecular details of suchphenomenon are lacking. In this paper, we investigate the interactions between thosenatural herbicides and artificial bilayers mimicking the plant plasma membrane. First,their ability to affect lipid order and fluidity is evaluated by means of fluorescencemeasurements. It appears that sorgoleone has a clear ordering effect on lipid bilayers,while nonanoic acid and sarmentine induce no or little change to these parameters.Then, a thermodynamic characterization of interactions of each compound with lipidvesicles is obtained with isothermal titration calorimetry, and their respective affinity forbilayers is found to be ranked as follows: sorgoleone > sarmentine > nonanoic acid.Finally, molecular dynamics simulations give molecular details about the location of eachcompound within a lipid bilayer and confirm the rigidifying effect of sorgoleone. Dataalso suggest that mismatch in alkyl chain length between nonanoic acid or sarmentineand lipid hydrophobic tails could be responsible for bilayer destabilization. Results arediscussed regarding their implications for the phytotoxicity of these compounds.

Keywords: plasma membrane, lipid, natural herbicides, phytotoxicity, allelopathy, sorgoleone, sarmentine,nonanoic acid

INTRODUCTION

Weeds management is a key point for an efficient agriculture and nowadays it mainly relies onthe use of herbicides. Unfortunately, traditional herbicides are becoming ineffective because of theevolution of resistance, and there are some public concern related to the potential environmentaland health impact of pesticides. In this context, identification of natural phytotoxic compounds

Abbreviations: DPPC, 1,2-Dipalmitoyl-sn-glycero-3-phosphocholine; ITC, isothermal titration calorimetry; LUV, largeunilamellar vesicle; MD, molecular dynamics; MLV, multilamellar vesicle; PLPC, 1-palmitoyl-2-linoleoyl-sn-glycero-3-phosphocholine.

Frontiers in Plant Science | www.frontiersin.org 1 March 2019 | Volume 10 | Article 329

fpls-10-00329 March 14, 2019 Time: 16:28 # 2

Lebecque et al. Natural Herbicides – Lipid Bilayers Interactions

with new modes of action is a promising way to reach amore sustainable agriculture. Much effort has been made in thisdirection, which enabled isolation and identification of numerousmolecules with an herbicidal activity (Dayan and Duke, 2014).However, a comprehensive understanding of the mechanismsunderlying the phytotoxicity of such molecules is mandatoryprior to their safe use on the field. In this paper, three naturallyoccurring compounds, namely nonanoic acid, sarmentine andsorgoleone were considered.

Nonanoic acid (Figure 1A) (also known as pelargonic acid)is produced by geranium (Pelargonium spp.). It has been usedfor years as a contact herbicide quickly leading to necroticlesions on aerial parts of the target plants. A comparison of thephytotoxicity of various fatty acids as a function of their carbonchain length demonstrated that fatty acids with C9 to C11 chainlengths where more active than fatty acids with shorter or longercarbon chains (Fukuda et al., 2004). According to Lederer et al.(2004), their phytotoxicity is due to fatty acids’ ability to penetrateand destabilize plasma membranes. However, there is no studyaccurately investigating the interactions between nonanoic acidand biological membranes.

Sarmentine (Figure 1B), a recently isolated amide from longpepper (Piper longum L.) fruit, has herbicidal activity on avariety of plant species (Huang et al., 2010). Plants treated withsarmentine develop very similar symptoms to those observed onplants treated with decanoic acid. The authors suggested thatsarmentine phytotoxicity is primarily exerted through membranedestabilization related to its alkyl chain length. More recentwork demonstrated that sarmentine is likely to have multiple

FIGURE 1 | Molecular structures of (A) nonanoic acid, (B) sarmentine, and(C) sorgoleone.

mechanisms of action, including inhibition of photosystem II andinhibition of enoyl-ACP reductase (Dayan et al., 2015). At thesame time, it was proven that sarmentine induces destabilizationof plasma membrane from cotyledon disks assays (Dayan et al.,2015). This effect was also observed with nonanoic acid, againsuggesting that both molecules share at least one mode ofaction. Sarmentine was, however, much more potent in inducingdestabilization of membrane integrity (Dayan et al., 2015).

Sorgoleone (Figure 1C) is an allelochemical compoundexuded from Sorghum bicolor (L.) Moench roots (Einhelligand Souza, 1992; Dayan et al., 2010). This molecule inhibitsphotosynthesis by interrupting the photosynthetic electrontransport in isolated chloroplasts (Nimbal et al., 1996; Gonzalezet al., 1997). However, sorgoleone lipophilicity prevents itstranslocation from roots to foliage and thus it is not able toreach its putative site of action (Hejl and Koster, 2004; Dayanet al., 2009). Moreover, even when it is directly applied on thefoliage, sorgoleone is ineffective on photosynthesis of matureplants (Dayan et al., 2009). In planta inhibition of photosynthesisseems to only occur in young seedlings. Hence, it is supposedthat sorgoleone exhibits its toxicity toward older plants throughother modes of action. For instance, Hejl and Koster observedthat root H+-ATPase is inhibited by sorgoleone (Hejl andKoster, 2004). They suggested that the resulting solute and wateruptake alteration could contribute to the phytotoxicity of theherbicidal molecule.

Nonanoic acid, sarmentine and sorgoleone have a broadspectrum toxicity (Einhellig and Souza, 1992; Dayan et al., 2012,2015) and share structural similarities: their molecular weightis low (<400 g/mol) and they all possess highly hydrophobicalkyl tails and somewhat polar heads. Given these commonfeatures, we can make the assumption that they could interactwith lipid membranes and that their efficacy might be linkedto this property. Concerning nonanoic acid and sarmentine,such interactions with cellular membranes are thought to be atleast partly responsible for their phytotoxicity, but the molecularunderstanding and details are lacking. The putative interactionsoccurring between sorgoleone and lipid bilayers is also evidentsince its target site in photosystem II is embedded in the thylakoidmembrane of chloroplasts. Due to its chemical structure andits low mobility within plant tissues supposedly due to a highlipophilicity, investigating the behavior of sorgoleone in thepresence of lipid membranes could bring valuable informationabout its mechanism of action.

Biological membranes are very complex structures composedof numerous lipid species forming a bilayer in which a varietyof proteins are embedded. Because of this complexity, modelmembranes are generally used to mimick biological membranes.Such model membranes are made of one or a few representativelipid species forming structures such as monolayers, bilayers orvesicles that share common features with biological membranes(Deleu et al., 2014). In this work, interactions betweeneach of the above herbicidal molecules and plant modelmembranes are explored. At first, the effect of the herbicidalcompounds on the lipid bilayer order and fluidity is studiedat various temperatures with fluorescent measurements. Then,a thermodynamic characterization of the interactions occurring

Frontiers in Plant Science | www.frontiersin.org 2 March 2019 | Volume 10 | Article 329

fpls-10-00329 March 14, 2019 Time: 16:28 # 3

Lebecque et al. Natural Herbicides – Lipid Bilayers Interactions

between the chemicals under investigation and lipid vesicles iscarried out by using ITC. Finally, molecular mechanisms of theseinteractions are explored by performing MD simulations.

MATERIALS AND METHODS

Chemicals1,2-Dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) andPLPC were purchased from Avanti Polar Lipids, Inc. Sarmentinewas purchased from Toronto Research Chemicals Inc. Allother chemicals, including nonanoic acid, laurdan and DPHwere purchased from Sigma-Aldrich. Sorgoleone was extractedfrom Sorghum bicolor (L.) Moench as described previously(Dayan et al., 2009).

Liposome PreparationsFor laurdan generalized polarization experiments, MLVsprepared as follows were used. A small volume of DPPCdissolved into chloroform-methanol (2:1) was either put aloneor mixed with small amounts of nonanoic acid, sarmentine orsorgoleone (dissolved into 2:1 chloroform-methanol) into around bottom tube to reach a lipid:herbicide molar ratio of 5:1.Solvent was then evaporated under a gentle stream of nitrogen.The tube was kept overnight under vacuum to remove solventtraces. The resulting film was then hydrated with 10 mM TRIS -HCl buffer at pH 7 prepared from Milli-Q water. The tube wasmaintained at a temperature (∼51◦C) well above the transitionphase temperature of the lipid for at least 1 h and vortexedfor 1–2 min every 10 min. Thereafter, the MLV suspensionunderwent 5 freeze-thaw cycles.

For DPH anisotropy measurements, pure DPPC LUVs wereprepared by following the same procedure followed by 11passages through polycarbonate filters with a pore diameter of100 nm at a temperature (∼51◦C) well above the transition phasetemperature of the lipid.

For ITC, LUVs were used. Small amounts of lipids (PLPC)were dissolved into chloroform-methanol (2:1) in a round-bottom flask. A rotary evaporator was used to remove solventunder low pressure and the flask was then kept overnightunder vacuum to remove solvent traces. The lipid film was thenhydrated with 10 mM TRIS - HCl buffer at pH 7 preparedfrom Milli-Q water. The flask was maintained at a temperature(∼37◦C) well above the transition phase temperature of the lipidfor at least 1 h and vortexed for 1–2 min every 10 min. Thereafter,the MLV suspension underwent 5 freeze-thaw cycles. In order toget LUVs, MLV suspension was then extruded 11 times throughpolycarbonate filters with a pore diameter of 100 nm.

Laurdan Generalized PolarizationFluorescence of laurdan in DPPC (either pure or mixed withone of the studied molecules in a lipid:herbicide molar ratio of5:1) MLVs was monitored at various temperatures with a PerkinElmer LS50B fluorescence spectrometer. Lipid concentration was50 µM and 1 µL of a stock solution of laurdan in DMSO wasadded to 3 ml of MLV suspension to reach a lipid:probe molarratio of 200:1. A corresponding blank was prepared by adding

1 µL of pure DMSO instead of the laurdan stock solution to 3 mlof MLV suspension from the same batch. Both preparations werekept at a temperature (∼51◦C) well above the phase transitiontemperature of DPPC for 1 h in the dark to let the probepartition into the MLVs. Samples were then placed in 10 mmpath length quartz cuvettes under continuous stirring whilethe cuvette holder was thermostated with a circulating bath.Samples were equilibrated at the initial temperature of 31◦Cfor 15 min before the first measurements were performed. Forthe measurements, excitation wavelength was set to 360 nm,and at least 12 measurements of emission intensities at 440and 490 nm were recorded and averaged for each sample ateach temperature. An emission spectrum from 400 to 600 nmwas also recorded for each sample-temperature combination.Excitation slit width was set to 2.5 nm, while emission slit widthranged from 4 to 6.5 nm. When all measurements at a giventemperature were recorded, temperature was increased by 4, 3,2, or 1 degree(s) and samples were equilibrated for 10 min oncethe new temperature was reached. Generalized polarization (GP)of laurdan was then calculated according to (Parasassi et al., 1991;Jay and Hamilton, 2017):

GP =I440 − I490

I440 + I490(1)

where I440 and I490 are the blank-subtracted emission intensitiesat 440 and 490 nm, respectively. All experiments werecarried out at least three times, each time with liposomesprepared fresh daily.

DPH AnisotropyFluorescence anisotropy measurements of DPH in DPPC LUVseither in the gel phase (at 35◦C) or in the liquid phase (at 50◦C)were performed with a Perkin Elmer LS50B fluorescencespectrometer equipped with polarizers. To prepare the samples,stock solutions of nonanoic acid, sarmentine or sorgoleonedissolved into DMSO at various concentrations were prepared.Four µL of each of these solutions were added to aliquots (3 mL)of 50 µM DPPC LUV suspension to reach final concentrationsof 10, 25, 50, and 100 µM of natural herbicide. For pure DPPCsample, 4 µL of DMSO were added to an aliquot of 50 µLDPPC LUV suspension. One µL of a stock solution of DPHdissolved into DMSO was then added to each sample to reacha lipid:probe molar ratio of 200:1. Corresponding blanks wereprepared by adding 1 µL of pure DMSO instead of the DPH stocksolution. Each preparation was kept at a temperature (∼51◦C)well above the phase transition temperature of DPPC for 1 h inthe dark to let the probe and the molecules partition into theLUVs. Samples were then placed in 10 mm path length quartzcuvettes under continuous stirring while the cuvette holder wasthermostated with a circulating bath. Samples were equilibratedat the desired temperature for 15 min before the measurementswere performed. For the measurements, excitation and emissionwavelengths were set to 360 and 426 nm, respectively. Excitationslit width was set to 2.5 nm, while emission slit width rangedfrom 4.5 to 12 nm. Both excitation and emission polarizers can beeither in a vertical (V) or a horizontal (H) orientation. For eachsample, at least 12 measurements were recorded and averaged for

Frontiers in Plant Science | www.frontiersin.org 3 March 2019 | Volume 10 | Article 329

fpls-10-00329 March 14, 2019 Time: 16:28 # 4

Lebecque et al. Natural Herbicides – Lipid Bilayers Interactions

each of the four possible combinations of polarizers orientations.The anisotropy parameter, r, was then calculated according toprevious papers (Harris et al., 2002; Lakowicz, 2006):

r =IVV − G× IVH

IVV + 2× G× IVH(2)

where IVV and IVH are the blank-subtracted emission intensitieswith the excitation and the emission polarizers oriented asindicated by the subscripts in this order (for instance, IVHcorresponds to the blank-subtracted emission intensity with theexcitation polarizer in a vertical orientation and the emissionpolarizer in a horizontal orientation) and G is a correction factorobtained as follows:

G =IHVIHH

(3)

where IHV and IHH are the blank-subtracted emission intensitieswith the excitation and the emission polarizers oriented asindicated by the subscripts in this order. All experimentswere carried out at least three times. Measurements at50◦C were carried out with liposomes prepared fresh andmeasurements at 35◦C were performed within 3 days followingliposomes preparation.

Isothermal Titration CalorimetryThe ITC measurements were performed with a VP-ITC fromMicrocal (Microcal Inc., Northampton, MA, United States). Thesample cell contained 1.4565 mL of a solution of nonanoic acid(500 µM), sarmentine (75 µM) or sorgoleone (75 µM) dissolvedin the same buffer as the LUV suspension. Reference cell wasfilled with mQ water. Small aliquots of LUV suspension wereadded to the sample cell with a software-controlled syringe. Thefirst injection was 2 µL and was not taken into account for datatreatment. It was followed by 28 successive additions of 10 µL,with an interval of 600 s. For nonanoic acid and sarmentine,a lipid concentration of 5 mM was used. As sorgoleone affinityfor lipid vesicles was higher than the other compounds tested,lipid concentration was lowered to 1 mM for ITC experimentsinvolving this molecule. For data treatment, Origin 7.0 softwarewas used as previously described (Heerklotz and Seelig, 2000;Zakanda et al., 2012) in order to fit the cumulative heats ofbinding to the equation:

i∑k=1

δhk = 1Hw→ bD VcellC0

AKC0

L1+ KC0

L(4)

where δhk is the heat produced after each injection, 1HDw → b is

the difference in molar enthalpy originating from the transfer ofthe herbicidal molecules from the aqueous phase to the bilayermembrane, Vcell is the volume of the sample cell, C0

A andC0

L are the total herbicide and lipid concentration, respectively,in the cell after i injections and K is the partition constant.Temperature was maintained at 26◦C during the course ofthe experiments. Data analysis was done using Origin software(version 7.0) from Microcal.

Molecular DynamicsSimulations have been performed with GROMACS 4.6.7 andthe united atom GROMOS 53a6 force field (Oostenbrink et al.,2004) with three replicates. Topologies of the natural herbicideshave been manually refined from Automatic Topology Builder’sresults (Malde et al., 2011). A PLPC topology derived from BergerLipids force field (Berger et al., 1997) and developed by PeterTieleman’s group (Bachar et al., 2004) was used. It is availablefor download on its website (Tieleman, 2018). Bilayers containing140 PLPC molecules and 28 herbicide (nonanoic acid, sarmentineor sorgoleone) molecules (lipid:herbicide molar ratio of 5:1) weregenerated and hydrated by using Memgen (Knight and Hub,2015). Each system was solvated with SPC water (Hermans et al.,1984). Nonanoic acid being unprotonated in our simulations,28 K+ ions (Gurtovenko and Vattulainen, 2008; Cordomí et al.,2009) were added to systems containing nonanoic acid moleculesto get an overall charge of zero. All systems firstly underwent a100 ps NVT equilibration followed by a 1 ns NPT equilibrationduring which herbicides were under position restraints. 250 nsproduction runs were then performed from which the first 150 nswere considered as equilibration time and discarded for dataanalyses. For the production runs, temperature was maintainedto an average value of 298K by using the Nose-Hoover thermostat(Nosé, 1984; Hoover, 1985) with a τT = 0.5 ps. Semi-isotropicpressure (1 bar) was maintained by using the Parrinello-Rahmanbarostat (Parrinello and Rahman, 1981) with a compressibilityof 4.5 × 10−5 bar−1 and τP = 2 ps. Electrostatic interactionswere treated by using the particle mesh Ewald (PME) method.A cut-off of 1nm was used for Van der Waals interactions. Bondlengths were maintained with the LINCS algorithm (Hess et al.,1997). Trajectories were analyzed with GROMACS tools as wellas with homemade scripts and were visually analyzed with VMD(Humphrey et al., 1996) and PYMOL (The PyMOL MolecularGraphics System) software packages.

RESULTS

Effect on Bilayer Order and Lipid PhaseTransition TemperatureLaurdan is a fluorescent probe that readily partitions at thehydrophilic/hydrophobic interface of bilayers (Bernik et al., 2001;De Vequi-Suplicy et al., 2006). Laurdan fluorescence propertiesdepend on the physical state of the probe environment. Wheninserted into a lipid bilayer in the gel phase, fluorescence intensityis maximal at an emission wavelength of about 440 nm. As thelipid bilayer undergoes transition to a liquid phase, the maximalfluorescence intensity will occur at higher emission wavelengths.This so-called “red shift” is imputed to an increased presence andmobility of water molecules around the fluorophore (Parasassiet al., 1991; De Vequi-Suplicy et al., 2006) which can be directlylinked to the lipid bilayer order (Harris et al., 2002). Laurdangeneralized polarization (GP) parameter has been developedin order to conveniently monitor these lipid order changes(Parasassi et al., 1991). Here, laurdan GP is used to study the effectof the natural herbicides on the phase transition temperature

Frontiers in Plant Science | www.frontiersin.org 4 March 2019 | Volume 10 | Article 329

fpls-10-00329 March 14, 2019 Time: 16:28 # 5

Lebecque et al. Natural Herbicides – Lipid Bilayers Interactions

(Tm) of DPPC MLVs as well as to investigate their effect on lipidorder at the hydrophilic/hydrophobic interface below, aroundand above the Tm of DPPC.

The temperature dependence of laurdan GP for MLVsmade of either pure DPPC or DPPC mixed with a naturalherbicide (lipid:herbicide molar ratio of 5:1) is shown inFigure 2. Nonanoic acid and sarmentine do not affect the phasetransition temperature of DPPC MLVs at this concentration.On the other hand, DPPC MLVs mixed with sorgoleoneexhibit a slightly increased Tm when compared to pure DPPC.However, the cooperativity of the phase transition does notseem to be affected by sorgoleone as no broadening of thecurve in this region is detected. In addition, we observe thatnonanoic acid and sarmentine do not induce any changein lipid order at any temperature as their respective curvessuperimpose the pure DPPC curve. For sorgoleone and DPPCmixed liposomes, laurdan GP values are higher both in the gelphase and in the liquid phase when compared to the otherMLV compositions. This likely indicates an ordering effect ofsorgoleone on DPPC bilayer.

Study of the Fluidity of the HydrophobicCore of Bilayer in the Presence of theNatural HerbicidesThe DPH is a hydrophobic fluorescent probe that spontaneouslyand almost exclusively inserts into the hydrocarbon chains regionof lipid bilayers (Kaiser and London, 1998) and is then widelyused to analyze the microviscosity of this region in artificialmembranes (Shinitsky and Barenholz, 1978).

As its excited state lifetime is sufficiently long, DPH canchange its orientation before fluorescence emission occurs. Theability of the probe to reorient within the hydrophobic core ofthe lipid bilayer is directly linked to the microviscosity/fluidityof the hydrocarbon chains of the bilayer (Lentz, 1989; Stottet al., 2008). When the exciting light is polarized, rotational

FIGURE 2 | Evolution of laurdan generalized polarization as a function oftemperature for DPPC MLVs (50 µM) in the absence or presence(lipid:herbicide molar ratio 5:1) of natural herbicides (2: pure DPPC, �:DPPC + nonanoic acid, ×: DPPC + sarmentine, N: DPPC + sorgoleone).

movements of the probe induce a change in polarizationfor the emission light (Lentz, 1993). The less fluid is thebilayer, the less rotational movements of DPH occur andthe more polarized is the emission light, which is translatedinto a high value of anisotropy parameter r (see Eq. 2,section “Materials and Methods”). Hence, this anisotropyparameter r has been monitored to investigate the effectof the natural herbicides at different concentrations on thebilayer fluidity.

The evolution of DPH anisotropy for DPPC LUVs (lipidconcentration = 50 µM) at 50◦C as a function of theconcentration of added natural herbicides is shown in Figure 3.At this temperature, DPPC LUVs are in the liquid phaseand thus exhibit low r-values. Nonanoic acid does not induceany significant change in DPH anisotropy in the range ofconcentrations tested here (10–100 µM, i.e., a lipid:herbicidemolar ratio ranging from 5:1 to 1:2). Sarmentine seems ableto slightly increase r, but only at the highest concentrationused in this study. Sorgoleone, on the other hand, inducesa sharp rise of DPH anisotropy even at concentration aslow as 10 µM. Albeit less marked, similar trends areobserved at 35◦C, when liposomes are in the gel phase (seeSupplementary Figure S1).

Partitioning Into Lipid VesiclesIsothermal titration calorimetry experiments were carried outto study the ability of the natural herbicides to partition intolipid bilayers and to thermodynamically characterize interactionsthat could occur.

Typical raw data of an ITC experiment for nonanoic acid(A), sarmentine (B) and sorgoleone (C) with PLPC LUVsare shown in Figure 4 (upper panels). The correspondingcumulative heats of binding (6δhi) are also plotted againstthe lipid concentration in the cell (C0

L) (lower panels). Fittingcurves are obtained from Eq. 4 and enabled determinationof the partition coefficient, K, as well as thermodynamic

FIGURE 3 | Evolution of DPH anisotropy for DPPC LUVs (50 µM) mixed withincreasing concentrations of natural herbicides at 50◦C (�: DPPC + nonanoicacid, ×: DPPC + sarmentine, N: DPPC + sorgoleone).

Frontiers in Plant Science | www.frontiersin.org 5 March 2019 | Volume 10 | Article 329

fpls-10-00329 March 14, 2019 Time: 16:28 # 6

Lebecque et al. Natural Herbicides – Lipid Bilayers Interactions

FIGURE 4 | Upper panels: raw data from ITC experiments. Each peak corresponds to a single injection of 10 µL of a PLPC LUV suspension [5 mM for (A) and (B),1 mM for (C)] into a solution containing (A) 500 µM nonanoic acid, (B) 75 µM sarmentine and (C) 75 µM sorgoleone at 26◦C. LUV suspension and herbicidessolutions were buffered at pH 7 with TRIS – HCl. Lower panels: cumulative heats of binding (6δhi) as a function of the lipid concentration in the cell (C0

L). Solid linecorresponds to the best fit using the Eq. 4.

parameters characterizing the interactions, as summarized inTable 1. ITC raw data display a gradual decrease of the positiveheat flow signal over the course of the successive injectionswhich is characteristic of a binding event. The thermodynamicparameters (Table 1) indicate that the binding is spontaneous(1G < 0), endothermic (1H > 0) and leads to a positivechange of entropy (1S > 0) for all three molecules. Theabsolute value of entropy change is much higher than the oneof enthalpy change, indicating that the binding is entropy-driven. This is probably due to the transfer of the hydrophobicchain of the herbicidal molecules from the water to thehydrophobic core of the liposomes. This is supported by thegood correlation observed between the number of carbon atomsin the alkyl chain of each compound and their respectivemagnitude of T1SD

w → b. Nonanoic acid has the lowest bindingaffinity among the studied compounds. Sarmentine partitioncoefficient is almost one order of magnitude higher thanthat of nonanoic acid while sorgoleone one is two orders ofmagnitude larger.

Interaction of the Natural HerbicidesWith the Lipid Bilayer and Their Effect onthe Lipid Order Parameter Studied byMolecular DynamicsMolecular dynamics simulations have been carried out in orderto study the molecular mechanisms underlying the interactionbetween the herbicidal molecules and a PLPC bilayer. Figure 5displays snapshots at the end of 250 ns simulations with28 molecules of (A) nonanoic acid, (B) sarmentine and (C)sorgoleone. From this figure, it can be seen that only sorgoleone,the largest molecule under investigation, is able to interact withevery part of the lipid hydrophobic tails, while nonanoic acidand sarmentine are too short to significantly interact with thecenter of the bilayer.

The average distance between the center of mass (COM) ofthe bilayer and various groups of the natural herbicides along theZ-axis is plotted on Figures 6A–C for nonanoic acid, sarmentineand sorgoleone, respectively. The average distance between the

TABLE 1 | Thermodynamic parameters characterizing the interactions of herbicidal molecules with PLPC LUVs.

Natural compound K (mM−1) 1HDw → b (kJ.mol−1) T1SD

w → b (kJ.mol−1) 1GDw → b (kJ.mol−1)

Nonanoic acid 0.36 ± 0.06 3.32 ± 0.67 27.85 ± 0.6 −24.53 ± 0.48

Sarmentine 1.69 ± 0.57 1.47 ± 0.61 29.77 ± 0.39 −28.3 ± 0.82

Sorgoleone 28.5 ± 8.33 1.79 ± 0.77 37.13 ± 0.68 −35.34 ± 0.8

Frontiers in Plant Science | www.frontiersin.org 6 March 2019 | Volume 10 | Article 329

fpls-10-00329 March 14, 2019 Time: 16:28 # 7

Lebecque et al. Natural Herbicides – Lipid Bilayers Interactions

FIGURE 5 | Snapshots after 250 ns simulations of a 140 molecules PLPCbilayer with (A) 28 nonanoic acid molecules, (B) 28 sarmentine molecules or(C) 28 sorgoleone molecules. For sake of clarity, water molecules and ions areomitted. Dark red: herbicidal molecules, green: carbon atoms, red: oxygenatoms, orange: phosphor atoms, blue: nitrogen atoms.

bilayer COM and the phosphorus (P) atoms of the lipid heads isalso plotted, as well as the average distance between the bilayerCOM and the glycerol backbone COM. From these figures, itappears that each molecule has its own insertion pattern. The lastcarbon atom of the alkyl chain of sorgoleone is the most deeplyburied into the hydrophobic core of the bilayer, while nonanoicacid has the shallowest one. This is not surprising, as sorgoleonehas the longest alkyl chain, followed by sarmentine and thennonanoic acid. Concerning the polar moiety of each molecule, it

is observed that acid group of nonanoic acid is located at the levelof the glycerol backbone. Its inability to penetrate deeper into thehydrophobic core of the bilayer is likely due to its negative netcharge. For sorgoleone and sarmentine, on the other hand, thepolar part is located slightly deeper into the bilayer. Finally, itcan be noted that the P atoms – bilayer COM distance and theglycerol backbone COM – bilayer COM distance are significantlyhigher for simulations with sorgoleone, which indicates that thismolecule induces a thickening of the bilayer.

This thickening effect of sorgoleone on the lipid bilayer is dueto its ordering properties illustrated on Figure 7, which shows theorder parameter of each carbon atom from the palmitoyl chainduring the last 100 ns of simulation for all systems. It is observedthat sorgoleone indeed induces a strong ordering effect for carbon1 to carbon 13 of the palmitoyl chain. Interestingly, nonanoicacid and sarmentine seem able to increase, but to a lesser extentthan sorgoleone, the order parameter of the first 7 carbon atomsonly. Similar trends, but less marked, are also observed for thelinoleoyl chain (data not shown). The lipid hydrophobic coreregion affected by the presence of the herbicide molecules seemsthus related to the penetration depth of the last carbon atomof their alkyl chain. Deeper it is, higher the rigidification effectis and larger the lipid region where the rigidification occurs is.Surprisingly, the presence of trans unsaturations in sarmentineand cis unsaturations in sorgoleone, which could have a fluidizingeffect on the packing of the lipid chains (Roach et al., 2004), seemsto not have an impact on the bilayer parameter order.

DISCUSSION

In this study, the interactions occurring between naturalherbicides and model membranes made of PC lipids wereinvestigated. Phospholipids are one of the major componentsof plant plasma membranes (PPM), and PC is their mostcommon phospholipid [(Murphy et al., 2011) and referencescited therein]. It is also reported that palmitic and linoleicacids are the main fatty acids found in PPM phospholipids[(Murphy et al., 2011) and references cited therein]. Hence, PLPCwas used in ITC experiments as well as for MD simulations.However, for fluorescence studies, DPPC was used in orderto explore the impact of the molecules under investigation onlipid bilayers in both gel and liquid phases as well as theireffect on phase transition temperature. It would not have beenpossible to study PLPC liposomes at their Tm and below withfluorescence measurements, as PLPC has a reported Tm of about−18.5◦C.(Morrow et al., 1996) DPPC Tm is around 41.5◦C,(Lewis and McElhaney, 1992) which is in good agreementwith the laurdan GP curve of pure DPPC liposomes from thiswork (Figure 3).

From the fluorescence assays, we demonstrated that nonanoicacid and sarmentine have little influence on the lipid order,both at the level of the bilayer hydrophilic/hydrophobic interface(Laurdan experiments) and at the level of the hydrophobic core(DPH experiments), and on the phase transition temperature atthe concentration tested here. In a recent work, C8 and C10fatty acids only had a minor effect on DPPC vesicles phase

Frontiers in Plant Science | www.frontiersin.org 7 March 2019 | Volume 10 | Article 329

fpls-10-00329 March 14, 2019 Time: 16:28 # 8

Lebecque et al. Natural Herbicides – Lipid Bilayers Interactions

FIGURE 6 | Evolution of the average distance between various atomic groups and the PLPC bilayer center of mass along the Z-axis (nm). Blue line: herbicide (A:nonanoic acid, B: sarmentine, and C: sorgoleone) COM, red line: herbicide (A: nonanoic acid, B: sarmentine, and C: sorgoleone) polar part (defined as shown on theinset), green line: last methyl group of herbicide (A: nonanoic acid, B: sarmentine, and C: sorgoleone) alkyl chain, violet line: PLPC phosphorus atoms, cyan line:PLPC glycerol groups.

Frontiers in Plant Science | www.frontiersin.org 8 March 2019 | Volume 10 | Article 329

fpls-10-00329 March 14, 2019 Time: 16:28 # 9

Lebecque et al. Natural Herbicides – Lipid Bilayers Interactions

FIGURE 7 | Order parameter for each carbon atom in the palmitoyl chain ofPLPC in the presence or absence of herbicidal molecules. Blue line: purePLPC, red line: PLPC + nonanoic acid (5:1 molar ratio), green line:PLPC + sarmentine (5:1 molar ratio), violet line: PLPC + sorgoleone (5:1molar ratio).

transition temperature at a lipid:fatty acid molar ratio of 1:1(Arouri et al., 2016). This is well in line with the absence ofsignificant effect observed for nonanoic acid (lipid:fatty acidratio of 5:1) in this study. Sarmentine was only able to inducea very slight decrease in membrane hydrophobic core fluidity,at the highest concentration (100 µM). Only sorgoleone haddramatic effects on model membranes properties. Its ability toincrease DPPC Tm and order, as well as the strong decreaseof DPPC hydrophobic core fluidity indicate that sorgoleone hasan overall rigidifying effect on lipids. Alkylresorcinols with astructure similar to sorgoleone were recently found to have suchimpact on lipid bilayers, (Zawilska and Cieslik-Boczula, 2017) inagreement with the data presented here.

We clearly demonstrated that the investigatedcompounds do interact with PLPC LUVs using ITC. Theirrespective affinity for such bilayers is ranked as follows:sorgoleone > sarmentine > nonanoic acid. This is meaningfulwhen interpreting fluorescence results: the weaker partitioningof nonanoic acid and sarmentine into liposomes partly explainswhy these molecules had almost no impact on the studiedfluorescence parameters.

With our MD approach, natural herbicides are embeddedin the lipid bilayer from the beginning of the simulations.MD data confirmed that sorgoleone induces an ordering anda thickening effect on the PLPC bilayer. Interestingly, it alsoshowed that nonanoic acid and sarmentine slightly increasethe order parameter of the upper part of lipids. This effectcould stem from an alkyl chain length mismatch between theherbicidal molecules and PLPC. In addition, MD gave interestingdetails about these interactions at the molecular level. The deeperlocation of sorgoleone within the bilayer could partially explainits higher impact on lipid properties when compared to nonanoic

acid, for instance. Moreover, sorgoleone polar head comprisingan alcohol, two carbonyl and one methoxy, could have a higherpropensity to form intermolecular electrostatic interactions andH-bonds with distinct lipid headgroups/glycerol residues. By thisway, sorgoleone molecules can act as an anchor condensingthe lipids at the level of their polar heads that gives rise to arigidification of the hydrocarbon region as well, as suggested forthe rigidifying effect of sphingomyelin (Shinitsky and Barenholz,1978). The presence of cis unsaturations at the end of thesorgoleone chain seems to not be able to break the anchoringeffect of the polar head.

From these considerations, it appears that nonanoic acid andsarmentine have lower potency to affect lipid bilayer propertiesthan sorgoleone. However, it should be reminded that modelmembranes used in this study were made of PC lipids only. Thislipid species is a major component of plant plasma membranes,but specific interactions of nonanoic acid or sarmentine withother lipid species cannot be ruled out. Nonetheless, nonanoicacid and sarmentine spontaneously bind to PLPC liposomes inITC assays. Hence, it is possible that the potential target onwhich they exert their toxicity is embedded in membranes. If bothmolecules have a common target located in membrane, the higheraffinity of sarmentine for lipids might at least partly explain itshigher toxicity compared to nonanoic acid (Dayan et al., 2015).However, lipid affinity should not be the only factor for nonanoicacid toxicity. Indeed, the phytotoxicity of fatty acids is maximalwhen they are made of 9–11 carbon atoms (Fukuda et al., 2004)while the affinity of fatty acids for lipid vesicles increases whenincreasing the number of carbon atoms (Pernille et al., 2001),suggesting that toxicity and lipid affinity are not fully related.

From there, there are two options: (1) fatty acids phytotoxicityis not related to interactions with lipid bilayers or (2) an unknownfeature of their mode of action is lowered when alkyl chain lengthis increased. This putative feature might be an alkyl chain lengthmismatch between the fatty acids and the membrane lipids. Ithas been reported that the degree of mismatch between alkylchain length of lysophospholipids and membrane lipid chainlength is proportional to the bilayer perturbation caused bylysophospholipids (Henriksen et al., 2010). Hence, long chainfatty acids may have a high affinity for lipid bilayers but alow chain length mismatch with them, while short chain fattyacids have a high degree of mismatch with lipid bilayers buta low affinity for them. According to this assumption, middlechain length fatty acids would benefit from intermediate levelsof affinity and mismatch, with C9 to C11 presenting optimalcombinations of these two factors and thus exhibiting the highestability for lipid bilayer perturbation and eventually the highestphytotoxicity. Effects of such alkyl chain length mismatches weresuggested by our MD simulations but could probably not beefficiently probed with our fluorescent assays. Techniques thatenable lipid order measurements with a resolution of a singlemethyl group such as 2H NMR could be used to detect suchordering-disordering effect due to alkyl chain length mismatch.

The ordering effect that sorgoleone exerts on lipid bilayers isexpected to have some consequences on membrane functions. Ithas been suggested that small molecules such as phytochemicals(Ingólfsson et al., 2014) or anesthetics (Cantor, 1997, 2003;

Frontiers in Plant Science | www.frontiersin.org 9 March 2019 | Volume 10 | Article 329

fpls-10-00329 March 14, 2019 Time: 16:28 # 10

Lebecque et al. Natural Herbicides – Lipid Bilayers Interactions

Heimburg and Jackson, 2007) could modulate the activity ofmembrane-embedded proteins through global perturbations ofthe lipid environment. Interestingly, Hejl and Koster haveshown that sorgoleone inhibits the H+-ATPase activity (Hejland Koster, 2004). This transmembrane enzyme is responsiblefor the electrochemical gradient that enables metabolites uptakeby the plant cells. A decreased activity of this crucial protein,potentially due to changes in lipid environment properties, wouldthus have a major impact on cell viability. The relationshipbetween H+-ATPase activity and plant lipid environment hasbeen recently reviewed (Morales-Cedillo et al., 2015). It isreported that modifications in membrane lipid composition arerelated to alterations in enzymatic activity. For instance, thepresence of several sterols, molecules known for their orderingeffect, (Schuler et al., 1991; Bernsdorff and Winter, 2003)induces an inhibition of H+-ATPase. Albeit no straightforwardcorrelation between membrane fluidity and H+-ATPase activityhas been established, this tends to indicate that a link couldexist between the ability to alter lipid bilayer properties andthe modulation of proton-pumping activity could exist. Furtherresearch is, however, needed in order to understand how thelipid environment and its modification by molecules such assorgoleone could impact this enzyme.

AUTHOR CONTRIBUTIONS

MD and M-LF are the project leaders and contributed equallyto the work. SL performed the experiments and the overall data

analysis and wrote the manuscript. FD and LL provided technicaladvice. All authors revised the manuscript.

FUNDING

Partial computational resources have been provided by theConsortium des Équipements de Calcul Intensif (CÉCI), fundedby the Fonds de la Recherche Scientifique de Belgique (F.R.S.-FNRS) under Grant No. 2.5020.11. SL was funded by theCellule d’Appui à la Recherche et à l’Enseignement (CARE)AgricultureIsLife (University of Liège). MD and LL are SeniorResearch Associates for the Fonds National de la RechercheScientifique (FRS-FNRS).

ACKNOWLEDGMENTS

The authors thank the FNRS (PDR Grant T.1003.14; CDRJ.0014.08 and J.0086.18), and the University of Liège (Action deRecherche Concertée-Project FIELD) for financial support.

SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found onlineat: https://www.frontiersin.org/articles/10.3389/fpls.2019.00329/full#supplementary-material

REFERENCESArouri, A., Lauritsen, K. E., Nielsen, H. L., and Mouritsen, O. G. (2016).

Effect of fatty acids on the permeability barrier of model and biologicalmembranes. Chem. Phys. Lipids 200, 139–146. doi: 10.1016/j.chemphyslip.2016.10.001

Bachar, M., Brunelle, P., Tieleman, D. P., and Rauk, A. (2004). Molecular dynamicssimulation of a polyunsaturated lipid bilayer susceptible to lipid peroxidation.J. Phys. Chem. B 108, 7170–7179. doi: 10.1021/jp036981u

Berger, O., Edholm, O., and Jähnig, F. (1997). Molecular dynamics simulationsof a fluid bilayer of dipalmitoylphosphatidylcholine at full hydration, constantpressure, and constant temperature. Biophys. J. 72, 2002–2013. doi: 10.1016/S0006-3495(97)78845-3

Bernik, D. L., Zubiri, D., Tymczyszyn, E., and Disalvo, E. A. (2001). Polarity andpacking at the carbonyl and phosphate regions of lipid bilayers. Langmuir 17,6438–6442. doi: 10.1021/la010570s

Bernsdorff, C., and Winter, R. (2003). Differential properties of the sterolscholesterol, ergosterol, β-sitosterol, trans -7-dehydrocholesterol, stigmasteroland lanosterol on DPPC bilayer order. J. Phys. Chem. B 107, 10658–10664.doi: 10.1021/jp034922a.

Cantor, R. S. (1997). The lateral pressure profile in membranes: a physicalmechanism of general anesthesia. Biochemistry 36, 2339–2344. doi: 10.1021/bi9627323

Cantor, R. S. (2003). Receptor desensitization by neurotransmitters in membranes:?are neurotransmitters the endogenous anesthetics? Biochemistry 42,11891–11897. doi: 10.1021/bi034534z

Cordomí, A., Edholm, O., and Perez, J. J. (2009). Effect of force field parameters onsodium and potassium ion binding to dipalmitoyl phosphatidylcholine bilayers.J. Chem. Theory Comput. 5, 2125–2134. doi: 10.1021/ct9000763

Dayan, F. E., and Duke, S. O. (2014). Natural compounds as next-generationherbicides. Plant Physiol. 166, 1090–1105. doi: 10.1104/pp.114.239061

Dayan, F. E., Howell, J., and Weidenhamer, J. D. (2009). Dynamic root exudationof sorgoleone and its in planta mechanism of action. J. Exp. Bot. 60, 2107–2117.doi: 10.1093/jxb/erp082

Dayan, F. E., Owens, D. K., and Duke, S. O. (2012). Rationale for a natural productsapproach to herbicide discovery. Pest Manag. Sci. 68, 519–528. doi: 10.1002/ps.2332

Dayan, F. E., Owens, D. K., Watson, S. B., Asolkar, R. N., and Boddy, L. G. (2015).Sarmentine, a natural herbicide from Piper species with multiple herbicidemechanisms of action. Front. Plant Sci. 6:222. doi: 10.3389/fpls.2015.00222

Dayan, F. E., Rimando, A. M., Pan, Z., Baerson, S. R., Gimsing, A. L., andDuke, S. O. (2010). Sorgoleone. Phytochemistry 71, 1032–1039. doi: 10.1016/j.phytochem.2010.03.011

De Vequi-Suplicy, C. C., Benatti, C. R., and Lamy, M. T. (2006). Laurdan influid bilayers: position and structural sensitivity. J. Fluoresc. 16, 431–439. doi:10.1007/s10895-005-0059-3

Deleu, M., Crowet, J. -M., Nasir, M. N., and Lins, L. (2014). Complementarybiophysical tools to investigate lipid specificity in the interaction betweenbioactive molecules and the plasma membrane: a review. Biochim. Biophys. Acta1838, 3171–3190. doi: 10.1016/j.bbamem.2014.08.023

Einhellig, F. A., and Souza, I. F. (1992). Phytotoxicity of sorgoleone found in grainsorghum root exudates. J. Chem. Ecol. 18, 1–11. doi: 10.1007/BF00997160

Fukuda, M., Tsujino, Y., Fujimori, T., Wakabayashi, K., and Böger, P. (2004).Phytotoxic activity of middle-chain fatty acids I: effects on cell constituents.Pestic. Biochem. Physiol. 80, 143–150. doi: 10.1016/j.pestbp.2004.06.011

Gonzalez, V. M., Kazimir, J., Nimbal, C., Weston, L. A., and Cheniae, G. M. (1997).Inhibition of a photosystem II electron transfer reaction by the natural productsorgoleone. J. Agric. Food Chem. 45, 1415–1421. doi: 10.1021/jf960733w

Gurtovenko, A. A., and Vattulainen, I. (2008). Effect of NaCl and KCl onphosphatidylcholine and phosphatidylethanolamine lipid membranes: insightfrom atomic-scale simulations for understanding salt-induced effects in theplasma membrane. J. Phys. Chem. B 112, 1953–1962. doi: 10.1021/jp0750708

Frontiers in Plant Science | www.frontiersin.org 10 March 2019 | Volume 10 | Article 329

fpls-10-00329 March 14, 2019 Time: 16:28 # 11

Lebecque et al. Natural Herbicides – Lipid Bilayers Interactions

Harris, F. M., Best, K. B., and Bell, J. D. (2002). Use of laurdan fluorescence intensityand polarization to distinguish between changes in membrane fluidity andphospholipid order. Biochim. Biophys. Acta 1565, 123–128. doi: 10.1016/S0005-2736(02)00514-X

Heerklotz, H., and Seelig, J. (2000). Titration calorimetry of surfactant–membranepartitioning and membrane solubilization. Biochim. Biophys. Acta 1508, 69–85doi: 10.1016/S0304-4157(00)00009-5

Heimburg, T., and Jackson, A. D. (2007). The thermodynamics of generalanesthesia. Biophys. J. 92, 3159–3165. doi: 10.1529/biophysj.106.099754

Hejl, A. M., and Koster, K. L. (2004). The allelochemical sorgoleone inhibits rootH+-ATPase and water uptake. J. Chem. Ecol. 30, 2181–2191. doi: 10.1023/B:JOEC.0000048782.87862.7f

Henriksen, J. R., Andresen, T. L., Feldborg, L. N., Duelund, L., and Ipsen, J. H.(2010). Understanding detergent effects on lipid membranes: a model study oflysolipids. Biophys. J. 98, 2199–2205. doi: 10.1016/j.bpj.2010.01.037

Hermans, J., Berendsen, H. J. C., Van Gunsteren, W. F., and Postma, J. P. M. (1984).A consistent empirical potential for water–protein interactions. Biopolymers 23,1513–1518. doi: 10.1002/bip.360230807

Hess, B., Bekker, H., Berendsen, H. J., and Fraaije, J. G. (1997). LINCS: a linearconstraint solver for molecular simulations. J. Comput. Chem. 18, 1463–1472.doi: 10.1002/(SICI)1096-987X(199709)18:12<1463::AID-JCC4>3.0.CO;2-H

Hoover, W. G. (1985). Canonical dynamics: equilibrium phase-space distributions.Phys. Rev. A 31, 1695–1697. doi: 10.1103/PhysRevA.31.1695

Huang, H., Morgan, C. M., Asolkar, R. N., Koivunen, M. E., and Marrone, P. G.(2010). Phytotoxicity of sarmentine isolated from long pepper ( Piper longum )fruit. J. Agric. Food Chem. 58, 9994–10000. doi: 10.1021/jf102087c

Humphrey, W., Dalke, A., and Schulten, K. (1996). VMD: visual moleculardynamics. J. Mol. Graph. 14, 33–38. doi: 10.1016/0263-7855(96)00018-5

Ingólfsson, H. I., Thakur, P., Herold, K. F., Hobart, E. A., Ramsey, N. B., Periole, X.,et al. (2014). Phytochemicals perturb membranes and promiscuously alterprotein function. ACS Chem. Biol. 9, 1788–1798. doi: 10.1021/cb500086e

Jay, A. G., and Hamilton, J. A. (2017). Disorder amidst membrane order:standardizing laurdan generalized polarization and membrane fluidity terms.J. Fluoresc. 27, 243–249. doi: 10.1007/s10895-016-1951-8

Kaiser, R. D., and London, E. (1998). Location of diphenylhexatriene (DPH)and its derivatives within membranes: comparison of different fluorescencequenching analyses of membrane depth. Biochemistry 37, 8180–8190. doi: 10.1021/bi980064a

Knight, C. J., and Hub, J. S. (2015). MemGen: a general web server for thesetup of lipid membrane simulation systems. Bioinformatics 31, 2897–2899.doi: 10.1093/bioinformatics/btv292

Lakowicz, J. R. (2006). Principles of Fluorescence Spectroscopy. 3rd edn. New York,NY: Springer. doi: 10.1007/978-0-387-46312-4

Lederer, B., Fujimori, T., Tsujino, Y., Wakabayashi, K., and Böger, P. (2004).Phytotoxic activity of middle-chain fatty acids II: peroxidation and membraneeffects. Pestic. Biochem. Physiol. 80, 151–156. doi: 10.1016/j.pestbp.2004.06.010.

Lentz, B. R. (1989). Membrane “fluidity” as detected by diphenylhexatriene probes.Chem. Phys. Lipids 50, 171–190. doi: 10.1016/0009-3084(89)90049-2

Lentz, B. R. (1993). Use of fluorescent probes to monitor molecular order andmotions within liposome bilayers. Chem. Phys. Lipids 64, 99–116. doi: 10.1016/0009-3084(93)90060-G

Lewis, R., and McElhaney, R. N. (1992). The Mesomorphic Phase Behavior of LipidBilayers. CRC Press: Boca Raton, FL.

Malde, A. K., Zuo, L., Breeze, M., Stroet, M., Poger, D., Nair, P. C., et al. (2011).An automated force field topology builder (ATB) and repository: version 1.0.J. Chem. Theory Comput. 7, 4026–4037. doi: 10.1021/ct200196m

Morales-Cedillo, F., González-Solís, A., Gutiérrez-Angoa, L., Cano-Ramírez, D. L.,and Gavilanes-Ruiz, M. (2015). Plant lipid environment and membraneenzymes: the case of the plasma membrane H+-ATPase. Plant Cell Rep. 34,617–629. doi: 10.1007/s00299-014-1735-z

Morrow, M. R., Davis, P. J., Jackman, C. S., and Keough, K. M. (1996).Thermal history alters cholesterol effect on transition of 1-palmitoyl-2-linoleoylphosphatidylcholine. Biophys. J. 71, 3207–3214. doi: 10.1016/S0006-3495(96)79514-0

Murphy, A. S., Schulz, B., and Peer, W. (2011).The Plant PlasmaMembrane. Berlin:Springer-Verlag doi: 10.1007/978-3-642-13431-9

Nimbal, C. I., Yerkes, C. N., Weston, L. A., and Weller, S. C. (1996). Herbicidalactivity and site of action of the natural product sorgoleone. Pestic. Biochem.Physiol. 54, 73–83. doi: 10.1006/pest.1996.0011

Nosé, S. (1984). A molecular dynamics method for simulations in the canonicalensemble. Mol. Phys. 52, 255–268. doi: 10.1080/00268978400101201

Oostenbrink, C., Villa, A., Mark, A. E., and Van Gunsteren, W. F. (2004).A biomolecular force field based on the free enthalpy of hydration and solvation:the GROMOS force-field parameter sets 53A5 and 53A6. J. Comput. Chem. 25,1656–1676. doi: 10.1002/jcc.20090

Parasassi, T., De Stasio, G., Ravagnan, G., Rusch, R. M., and Gratton, E. (1991).Quantitation of lipid phases in phospholipid vesicles by the generalizedpolarization of laurdan fluorescence. Biophys. J. 60, 179–189. doi: 10.1016/S0006-3495(91)82041-0

Parrinello, M., and Rahman, A. (1981). Polymorphic transitions in single crystals:a new molecular dynamics method. J. Appl. Phys. 52, 7182–7190. doi: 10.1063/1.328693

Pernille H., Davidsen, J., and Jørgensen, K. (2001). Lipid membrane partitioningof lysolipids and fatty acids: effects of membrane phase structure and detergentchain length§. J. Phys. Chem. B 105, 2649–2657. doi: 10.1021/jp003631o

Roach, C., Feller, S. E., Ward, J. A., Shaikh, S. R., Zerouga, M., and Stillwell, W.(2004). Comparison of cis and trans fatty acid containing phosphatidylcholineson membrane properties. Biochemistry 43, 6344–6351. doi: 10.1021/bi049917r

Schuler, I., Milon, A., Nakatani, Y., Ourisson, G., Albrecht, A. M., Benveniste, P.,et al. (1991). Differential effects of plant sterols on water permeability and onacyl chain ordering of soybean phosphatidylcholine bilayers. Proc. Natl. Acad.Sci. U.S.A. 88, 6926–6930. doi: 10.1073/pnas.88.16.6926

Shinitsky, M., and Barenholz, Y. (1978). Fluidity parameters of lipid regionsdetermined by fluorescence polarization. Biochim. Biophys. Acta 515, 367–394.doi: 10.1016/0304-4157(78)90010-2

Stott, B. M., Vu, M. P., McLemore, C. O., Lund, M. S., Gibbons, E., Brueseke, T. J.,et al. (2008). Use of fluorescence to determine the effects of cholesterol on lipidbehavior in sphingomyelin liposomes and erythrocyte membranes. J. Lipid Res.49, 1202–1215. doi: 10.1194/jlr.M700479-JLR200

Tieleman, D. P. (2018). Biocomputing Group University of Calgary. Available at:http://wcm.ucalgary.ca/tieleman/downloads [accessed May 3].

Zakanda, F. N., Lins, L., Nott, K., Paquot, M., Lelo, G. M., and Deleu, M.(2012). Interaction of hexadecylbetainate chloride with biological relevantlipids. Langmuir 28, 3524–3533. doi: 10.1021/la2040328

Zawilska, P., and Cieslik-Boczula, K. (2017). Laurdan emission study of thecholesterol-like effect of long-chain alkylresorcinols on the structure ofdipalmitoylphosphocholine and sphingomyelin membranes. Biophys. Chem.221, 1–9. doi: 10.1016/j.bpc.2016.11.004

Conflict of Interest Statement: The authors declare that the research wasconducted in the absence of any commercial or financial relationships that couldbe construed as a potential conflict of interest.

Copyright © 2019 Lebecque, Lins, Dayan, Fauconnier and Deleu. This is an open-access article distributed under the terms of the Creative Commons AttributionLicense (CC BY). The use, distribution or reproduction in other forums is permitted,provided the original author(s) and the copyright owner(s) are credited and that theoriginal publication in this journal is cited, in accordance with accepted academicpractice. No use, distribution or reproduction is permitted which does not complywith these terms.

Frontiers in Plant Science | www.frontiersin.org 11 March 2019 | Volume 10 | Article 329